You administer a system with three nonglobal zones. All three-zones (z1, z2, and z3) use a common resource pool. The pool is a fixed, two-CPU configuration. The default scheduler is FSS. Zones z1…

Question

Options

• AZones z1 and z2 will consume CPU resources until each zone consumes approximately 20% of
• BZones z1 and z2 will consume CPU resources until each zone consumes approximately 60% of
• CZones z1 and z2 will consume CPU resources until each zone consumes approximately 90% of
• DZones z1 and z2 will consume CPU resources until each zone consumes all the resources that

• A
  19% (11)
• B
  7% (4)
• C
  71% (41)
• D
  3% (2)

Explanation

Option C is correct because the Fair Share Scheduler (FSS) dynamically redistributes unused CPU entitlement: z3 holds 60 of 100 total shares but is only consuming ~10% of the pool, so the remaining 90% of CPU capacity flows to the zones that are demanding it - z1 and z2 together absorb that 90%, splitting it equally (~45% each) since their share counts are identical.

Option A is wrong because 20% per zone reflects only the guaranteed floor proportional to their 20 shares out of 100 - FSS doesn't cap zones at their floor when idle capacity exists. Option B is wrong because 60% is z3's proportional entitlement, not z1/z2's. Option D is wrong because z3 is actively running and consuming 10%, so z1 and z2 cannot take all resources - the scheduler protects z3's active demand.

Memory tip: Think of FSS shares as a "claim ticket" that only matters when the system is contested - when another zone goes quiet, its unclaimed CPU is redistributed to hungry zones in proportion to their shares. Shares define the split of available CPU, not a hard cap.

No community discussion yet for this question.